Ultralow-Power Electronic Trapping of Nanoparticles with Sub-10 nm

Sep 7, 2016 - We demonstrate nanogap electrodes for rapid, parallel, and ultralow-power trapping of nanoparticles. Our device pushes the limit of diel...
0 downloads 3 Views 4MB Size
This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Letter pubs.acs.org/NanoLett

Ultralow-Power Electronic Trapping of Nanoparticles with Sub-10 nm Gold Nanogap Electrodes Avijit Barik,†,‡ Xiaoshu Chen,† and Sang-Hyun Oh*,†,‡ †

Department of Electrical and Computer Engineering and ‡Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota 55455, United States S Supporting Information *

ABSTRACT: We demonstrate nanogap electrodes for rapid, parallel, and ultralow-power trapping of nanoparticles. Our device pushes the limit of dielectrophoresis by shrinking the separation between gold electrodes to sub-10 nm, thereby creating strong trapping forces at biases as low as the 100 mV ranges. Using high-throughput atomic layer lithography, we manufacture sub-10 nm gaps between 0.8 mm long gold electrodes and pattern them into individually addressable parallel electronic traps. Unlike pointlike junctions made by electron-beam lithography or larger micron-gap electrodes that are used for conventional dielectrophoresis, our sub-10 nm gold nanogap electrodes provide strong trapping forces over a mm-scale trapping zone. Importantly, our technology solves the key challenges associated with traditional dielectrophoresis experiments, such as high voltages that cause heat generation, bubble formation, and unwanted electrochemical reactions. The strongly enhanced fields around the nanogap induce particletransport speed exceeding 10 μm/s and enable the trapping of 30 nm polystyrene nanoparticles using an ultralow bias of 200 mV. We also demonstrate rapid electronic trapping of quantum dots and nanodiamond particles on arrays of parallel traps. Our sub10 nm gold nanogap electrodes can be combined with plasmonic sensors or nanophotonic circuitry, and their low-power electronic operation can potentially enable high-density integration on a chip as well as portable biosensing. KEYWORDS: Nanogap, particle trapping, atomic layer deposition, dielectrophoresis, nanophotonics, atomic layer lithography

D

long range at 100 mV range biases without the disadvantages of optical trapping or traditional microelectrode-based DEP methods. Because many important biological particles, such as protein molecules and viruses, and quantum emitters are on the order of 10 nm in size, it is desirable to use electrodes with equally small gap sizes between them to maximize trapping efficiency and overcome the particle’s thermal Brownian motion. For a given bias voltage, smaller gaps can create stronger electric field gradients and can enable DEP trapping of particles at lower voltages, avoiding unwanted surface electrochemical reactions, bubble formation, or heat generation, all of which have been major challenges for applying DEP for broader applications. The fabrication of electrodes with sub-10 nm gaps, however, presents technical challenges. Previous work relied on slow and expensive techniques such as electron-beam lithography to create a small number of nanoscale gaps in bow-tie-like structures17,18 or by reducing the distance between carbonnanotube-based electrodes.19 Even when small gaps were realized, gold (which is desirable for optical sensing and

ielectrophoresis (DEP) has been widely investigated as a scalable technique to trap and manipulate polarizable objects such as cells, nanoparticles, DNA molecules, and proteins using electric field gradients.1 Although DEP manipulation of particles and molecules has been performed extensively using microelectrodes or sharp tips,2−4 significant hurdles still exist such as heat and bubble generation and unwanted surface reactions due to the high voltages required. Optical trapping systems5−7 have also been extensively used, but such schemes are plagued by drawbacks such as photobleaching and dependence on the refractive index contrast between the object and its medium and are typically limited to trapping a small number of objects in a confined trapping zone. Recent work on plasmonic trapping8−14 showed that the laser power required for trapping can be reduced compared with that necessary for optical trapping, but the localized evanescent optical fields limit the trapping range and often require additional fluidic manipulation to overcome the slow diffusive transport of particles.15 In comparison, DEP experiments can be performed using a simple and inexpensive apparatus (i.e., electrodes and a function generator) in a highly parallel fashion.16 We show that if the separation between electrodes can be reduced below 10 nm, DEP trapping can be performed over a © 2016 American Chemical Society

Received: June 29, 2016 Revised: September 1, 2016 Published: September 7, 2016 6317

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters

Figure 1. Fabrication of nanogap electrodes. (a) Fabrication scheme using atomic layer lithography. An Al2O3 layer of desired thickness (i.e., gap size) is deposited using ALD on a patterned gold film. A second layer of gold is evaporated, such that the first and second metal layers are not in contact. The top gold layer is then peeled off using adhesive tape, exposing the Al2O3-filled nanogap between the two gold electrodes. (b) An array of nanogap electrodes of desirable length is patterned by photolithography and ion milling on a 1 cm long nanogap. (c) A microscope image of a device containing an array of 32 nanogap electrodes. The size of each contact pad is 500 μm × 400 μm. (d) Magnified image of three electrodes, where the gaps are located at the center of each gold line. (e,f) SEMs of gold electrodes separated by 9 nm Al2O3 layer. (f) Scale bar: 200 nm.

plasmonics) could not be used due to surface reactions.20 In addition, these pointlike junctions can probe only a small amount of sample volume in the vicinity of the gap, sustaining the diffusion problem.21,22 Instead of using electron-beam lithography,23 many groups have demonstrated alternative schemes to create nanogaps with high throughput.24−27 Among various options, atomic layer lithography has shown the ability to create gaps that are as narrow as 1 nm in width with side lengths as large as 1 cm, producing strong confinement and enhancement of optical fields.28−31 Here, we adopt this method to manufacture a wafer-scale array of nanogap electrodes and use them to trap nanoparticles and quantum emitters at ultralow voltages. DEP force arises from the difference in conductivity and dielectric permittivity between an electrically polarizable particle and its surrounding medium.1 The time-averaged DEP force acting on a spherical particle of radius R is given by ⃗ (ω) = πεmR3·Re(f (ω))∇|E|2 FDEP CM

the single-digit nanometer regime. Furthermore, to enable massively parallel trapping and probe large sample volumes, we produce a large array of 0.8 mm long sub-10 nm gold electrodes. Shrinking the gap to ultrasmall length scales helps to rapidly trap nanoparticles, quantum dots, and nanodiamonds while applying biases in the range of only 100 mV. Applying such low voltages for DEP experiments with nanoparticles is unprecedented and solves the challenges associated with traditional high-voltage DEP experiments. In particular, Joule heating in the ionic solution, which scales as (voltage)2/ (resistance), can be significantly reduced. The fabrication scheme (Figure 1a), which creates precisely controlled nanoscale gaps in metal films, is based on atomic layer lithography29 that combines atomic layer deposition (ALD) with standard photolithography and tape peeling. First, an alumina (Al2O3) layer of the desired gap width (in this case, 9 nm) was deposited on a patterned gold substrate using ALD. Next, a second layer of gold was deposited directionally so that the first and second metal layers are not in contact. The excess gold film on top of the first layer was then peeled off using adhesive tape, exposing the nanogap between the two gold electrodes. Finally, to create an electrically disconnected array of nanogaps along with contact pads for electrical connection, standard photolithography was used, followed by ion milling (Figure 1b). Using this technique, we have fabricated up to 32 nanogaps per device on our prototype chip and varied the length of the gap from 20 μm to 0.8 mm. Figure 1c shows a microscope image (Keyence Digital Microscope) of an array of 32 nanogaps, where the gaps are aligned at the center of each gold line (Figure 1d). Details of the fabrication scheme can be found in the Supporting Information. An added benefit of our atomic layer lithography method is that the resulting nanogap is filled with an Al2O3 insulator, which prevents unwanted

(1)

where |E| is the magnitude of the electric field, εm is the permittivity of the surrounding medium, and Re( f CM(ω)) is the real part of the frequency (ω) dependent Clausius− Mossotti (CM) factor. Eq 1 shows that one can boost the DEP force either by (1) increasing the voltage between the electrodes (and thus |E|), which is not desirable due to unwanted heat generation, bubble formation, and electrochemical reactions, or (2) by reducing the distance between electrodes to increase ∇|E|2. In particular, because the DEP force has dimensions of V2m−3, shrinking the characteristic lengths of the system can significantly increase the force at a given bias voltage.32 In this work, we take this approach to the extreme case of sub-10 nm scaling and explore DEP trapping in 6318

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters

Figure 2. Ultralow-voltage DEP manipulation of 190 nm polystyrene beads. (a) Comparison of electrostatic field maps between 1 μm gap and 10 nm gap gold electrodes. It is clear that the nanogap electrodes can produce highly enhanced local electric fields and facilitate DEP trapping. The applied bias is 1 V. (b) The dependence of the minimum trapping voltage as a function of distance from the gap, d, is predicted for the two different particle sizes that were used in the experiments. The CM factor was assumed to be 1 for the particles. (c) Fluorescent image of floating beads in solution with the 9 nm gap noted by the dotted line. (d) The beads were attracted to the gap in positive DEP (300 mV, 1 MHz) and repulsed in (e) negative DEP (300 mV, 10 MHz). The direction of the particle motion is noted by arrows. (f) With positive DEP turned back on, the beads were once again trapped, showing that DEP trapping is repeatable and reversible. (g) Uniform trapping of polystyrene beads observed along a 0.8 mm long nanogap. (c−f) Scale bars: 20 μm. False-color fluorescent images are used to represent the emission wavelength.

nm particle is estimated to be ∼0.4 pN, which can be overcome by the DEP force generated from a 10 nm gap structure but not by using 1 μm gap. In Figure 2b, we investigated how the minimum voltage required to trap particles (Vmin) varies as a function of distance (d, as defined in Figure 2a) from the nanogap (9 nm width, as used in experiments), which is necessary to overcome diffusion limitations. Vmin was calculated by equating the DEP force (eq 1) with the maximum thermal force (eq 2). The dependence was tested for two different diameters of particle (10 and 30 nm), and it was observed to vary as Vmin ∝ d1.5. For example, a 30 nm particle can be trapped 1 μm away from the gap by applying ∼1 V of DC voltage (the equivalent alternating current (AC) peak amplitude being 1.4 V). A larger particle size or application of higher voltage can potentially increase the trapping range even further. On the contrary, to capture a 30 nm particle using a 1 μm gap electrode, the trapping range only extends up to 70 nm while 1 V is applied. Thus, by shrinking the gap width to sub-10 nm, we can perform long-range particle manipulation. For this analysis, the effect of electrode polarization was ignored because the operating frequency range in our experimental setup (100 kHz−10 MHz) is higher than the charge relaxation frequency (f) of the suspending medium of conductivity 4 μS/cm (90 kHz; f = σm/2πεm, σm: medium conductivity and εm: medium permittivity).32,33 Using highfrequency AC voltage for our experiments also ensures the absence of any unwanted electro-osmotic effects. We used 190 nm diameter carboxylate-modified polystyrene beads (λex: 470 nm, λem: 525 nm, Bangs Laboratories) to demonstrate efficient DEP trapping and releasing with ultralow

migration of gold atoms across the gap upon electrical biasing. Scanning electron micrographs (SEMs) of gold electrodes separated by a 9 nm Al2O3 layer are shown in Figure 1e,f at different magnifications. To test the isolation of the dielectric layer, a current−voltage (IV) curve was measured. From −2 to +2 V, it showed the characteristic noise of the experimental setup (Figure S1a). We also measured the open circuit potential of the device (using an electrochemical workstation; CH Instruments) to be approximately 1.2 mV in water, which is negligible compared to the voltages used in our experiments (Figure S1b). In our fabrication scheme, there is a 9 nm thick Al2O3 layer over one electrode, which can be removed by chemical etching in buffered oxide etch (BOE) solutions if desired. The magnitude of the electrostatic field distribution around the gap was compared for two different gap widths (1 μm and 10 nm) by finite-element method (FEM) simulations using COMSOL Multiphysics (Figure 2a). A 1 V direct current (DC) voltage was applied between the two electrodes separated by the gap. The maximum DEP force (10 nm away from the electrode boundary) on a 10 nm particle that can be generated from a 1 μm gap structure is approximately 0.2 pN, about 2 orders of magnitude smaller than the 37 pN generated from a 10 nm gap. To trap a small particle suspended in a solution, it is necessary to overcome the force due to its thermal motion (Fth), which for a particle of radius R is given by Fth =

kBTR 2R

(2)

where kB is the Boltzmann constant and TR is the ambient temperature.1 Using eq 2, the maximum thermal force on a 10 6319

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters

Figure 3. Parallel trapping and sub-0.5 V DEP manipulation of 30 nm polystyrene beads. (a) Schematic of an array of nanogap electrodes, each of which are individually addressable for DEP trapping. (b) Fluorescent image showing trapped 190 nm beads across three nanogaps with lengths of 20 μm each. A 20× objective was used to image three traps within the field of view. (c,d) SEMs of the trapped polystyrene beads (190 nm) along the nanogap at different magnifications. (e) Fluorescent image of a trapped 30 nm polystyrene bead (circled) on a 9 nm gap at a minimum bias of peak amplitude 200 mV. (f−h) More beads were trapped as the amplitude was increased further due to positive DEP at a frequency of 1 MHz. (i) The frequency was switched to 10 MHz to repel the beads using negative DEP. (j) DEP trapping was reversible, and the beads can be trapped again as the frequency was switched back to 1 MHz. Scale bars: (b) 40 μm; (c) 2 μm; (d) 400 nm; and (e−j) 20 μm. False-color fluorescent images are used to represent the emission wavelength.

was applied (Figure S3). A video (4× real time) is included in the Supporting Information that shows the clear trapping and releasing of beads within seconds. The random motion of the particles due to diffusion followed by a deterministic movement due to DEP was quantified using a two-dimensional particle tracking software.34 A mean transport speed of 13.6 μm/sec was generated under the influence of DEP (details in the Supporting Information) toward the nanogap, perpendicular to the gap axis. Furthermore, this analysis also revealed that 190 nm particles could experience a trapping force from 5 to 10 μm away from the nanogap electrodes. Theoretically, these particles can be trapped from 6 μm away from a 30 nm gap junction while 1 V was applied, which falls within the range of experimental observation. We also found the minimum trapping voltage amplitude for 190 nm beads on 30 nm gaps to be approximately 600 mV (Figure S5). Because DEP is a gradient-dependent phenomenon, shrinking the gap width to 9 nm reduced the operating voltage to 300 mV (Figure 2d−f). A long-standing hurdle in DEP experiments has been to perform particle manipulation in conductive buffers without causing excessive heating. If high voltages (>5 V) are used, it is common to experience joule heating and bubble formation that limits the use of DEP in many applications.35 Here, we show trapping of 190 nm beads in 0.01× phosphate-buffered saline (PBS) of conductivity 182 μS/cm by applying a 1 V bias (Figure S6a). As we increase the conductivity further to 0.1× or 1× PBS, the relative polarizability of the polystyrene beads with respect to the surrounding medium becomes negative (Figure S6b), resulting in no trapping. However, even with increased

bias voltages (300 mV amplitude). The direction of the DEP force depends on the CM factor ( f CM(ω)), which is given by fCM (ω) =

εp*(ω) − εm*(ω) εp*(ω) + 2εm*(ω)

(3)

where ε*p (ω) and ε*m(ω) are the complex permittivities of the particle and the medium, respectively. Particles can either be attracted toward the nanogap due to positive DEP (Re( f CM(ω)) > 0) or get repelled due to negative DEP (Re( f CM(ω)) < 0). The frequency at which this transition happens is called the crossover frequency (f CM(ω) = 0). Fluorescent images (Figure 2c−f) show floating beads before DEP followed by reversible trapping by positive (1 MHz) and negative DEP (10 MHz). The crossover frequency of the polystyrene beads in a solution of conductivity 4 μS/cm (measured by B-771 LAQUAtwin; Horiba Scientific) was calculated to be 4 MHz (Figure S2), justifying the choice of 1 MHz for positive and 10 MHz for negative DEP. Such a clear demonstration of nanoparticle manipulation in an electric field created by sub-0.5 V bias is unprecedented. Furthermore, as the entire length of the gap acts as a DEP hotspot, the trapping process is very fast. Figure 2g shows the trapping of beads on a 9 nm gap along its entire 0.8 mm length. We have also performed extensive studies to manipulate beads on wider junctions (30 nm) before moving to sub-10 nm gaps. The results are included in the Supporting Information. The 190 nm diameter polystyrene beads were manipulated by positive and negative DEP while a bias of peak amplitude of 1 V 6320

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters

Figure 4. Sub-1 V DEP manipulation of quantum dots and nanodiamonds. (a) Time-lapse fluorescent images show trapped quantum dots along the nanogap, noted as a white dotted line in the first frame. The dots with core and shell size of 8 nm were trapped by positive DEP (1 MHz) while a minimum voltage of peak amplitude of 750 mV was applied. Trapped dots were released by simply turning off the bias, which is reversed by turning the bias back on. (b) Trapping of 40 nm diamond particles was first observed at an AC bias of peak amplitude 400 mV, which was further improved at higher amplitude of 600 mV. A 100 kHz frequency was used for positive and 10 MHz for negative DEP. (c) Time-lapse fluorescent images show instant release of a trapped ND particle from the nanogap, which diffuses away from the field of view within 10 s. (a−c) Scale bar: 10 μm. False-color fluorescent images are used to represent the emission wavelength.

conductivity of the solution (182 μS/cm as compared to 4 μS/ cm for water), our device can trap individual polystyrene nanoparticles and does not suffer from bubble generation. Our experimental setup for performing trapping across an array of individually addressable nanogaps is depicted in Figure 3a. Each gap can trap particles over its entire length (from micrometer to millimeter and centimeter depending on the design) upon application of an AC bias. A fluidic cell was made with a polydimethylsiloxane (PDMS) ring sealed with a coverslip on top while a broadband light source and CCD camera were used to observe the trapping events in real time. More details on the experimental setup can be found in the Supporting Information. Here, we demonstrate trapping of 190 nm polystyrene beads on an array of 20 μm long nanogaps (Figure 3b). Beads can be trapped either across multiple gaps at the same time or on a single gap by individually biasing electrodes. This capability can be useful for spatial and temporal control over nanoparticle manipulations. SEMs taken after trapping show the trapped beads along the entire gap at different magnifications (Figure 3c,d). Next, we demonstrate manipulation of 30 nm diameter polystyrene beads (λex: 470 nm, λem: 505 nm; Sigma-Aldrich) in Figure 3e−j. To avoid any formation of aggregates, carboxylatemodified beads were used in the experiments. The minimum voltage amplitude that could trap individual beads (circled in Figure 3e) was 200 mV. The bias was then gradually increased up to 450 mV and trapped increasingly more beads along the gap (Figure 3f−h) by positive DEP at a frequency of 1 MHz. A variance in the fluorescence intensity of individual beads is observed, which is likely due to (1) a variation in the number of fluorescent tags per beads, (2) a disparity in the bead size, or (3) aggregation of the beads. The frequency of the AC bias was

then changed to 10 MHz to repel the beads from the gap by negative DEP (Figure 3i). Most of the beads were repelled, but some of the beads adhered strongly to the surface and could not be removed. Finally, the frequency was switched back to 1 MHz, and the beads were trapped again (Figure 3j), demonstrating the reversible nature of particle manipulation. Joule heating is common in experiments involving high voltages. The increment in solution temperature (Ts) can be estimated using the Poisson’s equation with Joule heating as the energy source (k∇2Ts = −σ⟨E2⟩) and approximated as ΔTs ∼

σV 2 2k

(4)

where V is the amplitude of the AC signal, and σ and k are the electrical and thermal conductivity of the solution, respectively.36 The gradient in temperature rise within the system is a function of the power dissipation, which in turn depends on the applied voltage. To our knowledge, our experiments report the lowest operating bias used for nanoparticle manipulation, which is about 1 order of magnitude smaller than previously reported values using pointlike nanogap junctions.17−20 Lowering the amplitude by 10-fold reduces the temperature rise via Joule heating as well as reducing the power consumption by 100-fold. In addition, our long nanogap structure can probe a much larger trapping volume. For example, a 30 nm particle can be trapped 1 μm away from the gap by applying 1 V (Figure 2b). For pointlike nanogaps, it corresponds to 2.1 fL, assuming a hemispherical trapping volume. However, for an 800 μm long nanogap, the volume is ∼1000 times larger (1.26 pL), assuming a semicylindrical trapping volume. This mm-scale long trapping zone can further help to avoid diffusion-limited mass transport and in turn enable lower operating voltages−opening up 6321

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters

Figure 5. Assembly of nanodiamond microchips. (a) Schematic of individually addressable parallel nanogap traps. (b−d) Series of fluorescent images (gray scale for better representation) show the precise positioning of individual 40 nm diamond particles on each gap by positive DEP (1 V, 100 kHz). From the onset of applying bias (t1 = 0), it took 77 s to position two ND particles on two parallel traps. In this experiment, both nanogaps were biased simultaneously and imaged by a 50× objective. (e) Both ND particles were released instantly by simply switching the frequency to 10 MHz, and they diffused away from the field of view within seconds. (f) Finally, we turn on the positive DEP again by applying a higher amplitude of 1.5 V and observed nearly instant trapping (t2 = 5 s) of individual ND particles on the gap.

energy,9,25,40,41 our gold nanogaps can serve as a practical platform for plasmonics and nanophotonics. NDs with NV centers have been of great interest for applications in quantum information processing, sensing, and biological applications.6,42−44 To demonstrate the potential to realize ND-based microchips, we have made individually addressable parallel nanogaps (Figure 5a) to position single NDs on each gap and build an assembly of ND microchips. The length and spacing between the nanogaps can be adjusted depending on the application and in this work were 20 μm long and 80 μm apart, allowing us to fit two microchips within the field of view while using a 50× objective. Figure 5b−d demonstrate sequential trapping of ND particles on each microchip within 77 s while the gaps were biased simultaneously with 1 V peak amplitude and 100 kHz frequency. Trapped ND particles can be instantly released by switching the frequency to 10 MHz (Figure 5e) and reversibly trapped by simply switching the frequency back to 100 kHz. Figure 5f shows reversible trapping of ND particles within 5 s, which could be due to (1) application of higher voltage (1.5 V) or (2) the particles released in the previous step being in the vicinity of the traps. Although the trapping method used in this case can reliably position single NDs on each of the two adjacent nanogap traps, to make a larger array of ND microchips, a more-robust particle immobilization method could be implemented.15 This proof-of-concept experiment shows that our platform can make individually addressable microchips within the field of view and can position single quantum emitter on each of them to build sophisticated solid-state quantum optics devices. In conclusion, we have demonstrated ultralow-power dielectrophoresis using 0.8 mm long and sub-10 nm wide gold nanogap electrodes fabricated via atomic layer lithography. These devices show clear evidence of rapid, long-range particle manipulation (trap and release) at bias voltages as low as 200 mV, which is an unprecedented improvement over previously used DEP microelectrodes. This in turn solves the problems associated with high-voltage DEP, such as unwanted surface reactions, bubble formation, heat generation, and denaturation

avenues to use our devices as noninvasive electronic traps for biological particles such as cells, viruses, and nanovesicles. Rapid and reproducible localization of quantum emitters on prefabricated chips is highly desirable for experiments in nanooptics6,37,38 but is very difficult to accomplish due to their ultrasmall dimensions and Brownian motion in solutions. Here, we demonstrate the utility of our device by trapping core−shell CdSe quantum dots with a size of 8 nm (λem: 625 nm, Qdot 625 ITK Carboxyl Quantum Dots; Life Technologies) and 40 nm nanodiamond (ND) particles with 10−15 nitrogen-vacancy (NV) centers (λem: 637 nm; Adamas Nanotechnologies) on the gap while a bias of sub-1 V peak amplitude was applied (Figure 4). Both the quantum dots and ND particles are carboxylatemodified to ensure facile dispersion in aqueous solution. A series of fluorescent images were collected at 60 s time intervals to demonstrate trapping and releasing of quantum dots along the nanogap (Figure 4a). Trapping was observed at 750 mV and 1 MHz frequency. The quantum dots were not strongly adhered to the surface as they diffused away as soon as the bias was turned off. The dots could be trapped again once the bias was turned back on, demonstrating reversible trapping. We then performed the trapping experiments with the ND particles. Time-lapse fluorescent images (Figure 4b; taken at 50 s time intervals using λex: 540−553 nm) demonstrate the trapping of ND particles as the amplitude of the AC bias was varied from 400 to 600 mV at 100 kHz. NDs are dielectric particles that are polarized in the presence of a nonuniform electric field and are attracted toward the region of strongest field intensity by positive DEP. Similar to polystyrene beads, NDs can be repelled by negative DEP at higher frequencies, as shown in Figure 4c. A 10 MHz frequency was used to facilitate the instant release of a trapped ND particle from the gap, and it diffused away from the field of view within 10 s. A video (2× real time) to show trap and release of nanodiamond is included in the Supporting Information. Trapping quantum emitters by ultralow-power electronic nanogap electrodes is an important finding for building solid-state quantum optics devices.39 Furthermore, as nanoscale gap between noble metals such as gold are capable of extreme confinement of optical 6322

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters

program. Electron microscopy was performed at the Characterization Facility, which has received capital equipment funding from the NSF MRSEC. Computational modeling was carried out using software provided by the University of Minnesota Supercomputing Institute. The authors thank Jonah Shaver for his help with the optics setup, Daniel Klemme for helpful comments on the manuscript, Nathan Wittenberg for aiding in electrochemical measurements, and Shailabh Kumar for assistance with scanning electron microscopy.

of biomolecules. Also, as particles are trapped over an elongated nanogap, with this method it is possible to probe a much larger trapping volume as compared to pointlike nanoscale junctions. Further improvements to trap biomolecules in physiological solutions could include the integration of a convective-flowbased setup to dissipate heat or protecting the electrode surface with a passivation layer to avoid surface reactions in addition to application of low voltages. A fundamental limiting factor of surface-based biosensors is the slow diffusion of analytes into the sensing area.45 Based on our observations, it should be possible to design the channels in a way that all the analyte molecules within the reservoir can be concentrated on the gap, thus creating a “perfect sink”. Furthermore, gold nanogap electrodes are also capable of extreme confinement of optical energy, which can be used for concurrent plasmonic biosensing and surface-enhanced spectroscopy, potentially down to singlemolecule resolution. Finally, our electronic nanogaps can serve as an integration platform for applications in plasmonic sensing,46−48 nanophotonic circuits,38 single-molecule trapping,14 sub-10 nm precision positioning, and single-photon source arrays.38 Ultralow-power electronic operation of nanogap traps demonstrated herein shows their potential for highdensity on-chip integration as well as for portable sensing applications.





ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b02690. Additional details on experimental methods. Figures showing a sample current−voltage (IV) curve, opencircuit potential of the nanogap device, a Clausius− Mossotti factor plot for 190 nm polystyrene beads in water, trapping of 190 nm polystyrene beads on a 30 nm gap device, results from a two-dimensional particle tracking software used to track the trajectories of five different particles during diffusion and under the influence of the DEP force, time-lapse fluorescent images representing the voltage-dependence study of the 190 nm polystyrene beads on a 30 nm gap electrode, and the results from the trapping of 190 nm polystyrene beads in 0.01× PBS. (PDF) A movie showing the positive and negative DEP of individual polystyrene beads (190 nm). (AVI) A movie showing the positive and negative DEP of nanodiamonds (40 nm).(AVI)



REFERENCES

(1) Pohl, H. A. Dielectrophoresis: The Behavior of Neutral Matter in Nonuniform Electric Fields; Cambridge University Press: Cambridge, England, 1978. (2) Becker, F. F.; Wang, X.-B.; Huang, Y.; Pethig, R.; Vykoukal, J. V.; Gascoyne, P. R. C. Proc. Natl. Acad. Sci. U. S. A. 1995, 92, 860−864. (3) Hu, X.; Bessette, P.; Qian, J.; Meinhart, C.; Daugherty, P. S.; Soh, H. T. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 15757−15761. (4) Freedman, K. J.; Otto, L. M.; Ivanov, A. P.; Barik, A.; Oh, S.-H.; Edel, J. B. Nat. Commun. 2016, 7, 10217. (5) Grigorenko, A. N.; Roberts, N. W.; Dickinson, M. R.; Zhang, Y. Nat. Photonics 2008, 2, 365−370. (6) Geiselmann, M.; Juan, M. L.; Renger, J.; Say, J. M.; Brown, L. J.; de Abajo, F. J. G.; Koppens, F.; Quidant, R. Nat. Nanotechnol. 2013, 8, 175−179. (7) Yang, A. H. J.; Moore, S. D.; Schmidt, B. S.; Klug, M.; Lipson, M.; Erickson, D. Nature 2009, 457, 71−75. (8) Novotny, L.; Bian, R.; Xie, X. Phys. Rev. Lett. 1997, 79, 645−648. (9) Xu, H.; Käll, M. Phys. Rev. Lett. 2002, 89, 246802. (10) Juan, M. L.; Gordon, R.; Pang, Y.; Eftekhari, F.; Quidant, R. Nat. Phys. 2009, 5, 915−919. (11) Kang, J.-H.; Kim, K.; Ee, H.-S.; Lee, Y.-H.; Yoon, T.-Y.; Seo, M.K.; Park, H.-G. Nat. Commun. 2011, 2, 582. (12) Wang, K.; Schonbrun, E.; Steinvurzel, P.; Crozier, K. B. Nat. Commun. 2011, 2, 469. (13) Saleh, A. A. E.; Dionne, J. A. Nano Lett. 2012, 12, 5581−5586. (14) Pang, Y.; Gordon, R. Nano Lett. 2012, 12, 402−406. (15) Ndukaife, J. C.; Kildishev, A. V.; Nnanna, A. G. A.; Shalaev, V. M.; Wereley, S. T.; Boltasseva, A. Nat. Nanotechnol. 2016, 11, 53−59. (16) Barik, A.; Otto, L. M.; Yoo, D.; Jose, J.; Johnson, T. W.; Oh, S.H. Nano Lett. 2014, 14, 2006−2012. (17) Bezryadin, A.; Dekker, C.; Schmid, G. Appl. Phys. Lett. 1997, 71, 1273−1275. (18) Hölzel, R.; Calander, N.; Chiragwandi, Z.; Willander, M.; Bier, F. F. Phys. Rev. Lett. 2005, 95, 128102. (19) Tuukkanen, S.; Toppari, J. J.; Kuzyk, A.; Hirviniemi, L.; Hytönen, V. P.; Ihalainen, T.; Törmä, P. Nano Lett. 2006, 6, 1339− 1343. (20) Lesser-Rojas, L.; Ebbinghaus, P.; Vasan, G.; Chu, M.-L.; Erbe, A.; Chou, C.-F. Nano Lett. 2014, 14, 2242−2250. (21) Sheehan, P. E.; Whitman, L. J. Nano Lett. 2005, 5, 803−807. (22) Dahlin, A. B. Plasmonic Biosensors: An Integrated View of Refractometric Detection; IOS Press: Washington, DC, 2012. (23) Koh, A. L.; Fernandez-Dominguez, A. I.; McComb, D. W.; Maier, S. A.; Yang, J. K. W. Nano Lett. 2011, 11, 1323−1330. (24) Miyazaki, H. T.; Kurokawa, Y. Phys. Rev. Lett. 2006, 96, 097401. (25) Ward, D. R.; Grady, N. K.; Levin, C. S.; Halas, N. J.; Wu, Y.; Nordlander, P.; Natelson, D. Nano Lett. 2007, 7, 1396−1400. (26) Ciracì, C.; Hill, R. T.; Mock, J. J.; Urzhumov, Y.; FernandezDominguez, A. I.; Maier, S. A.; Pendry, J. B.; Chilkoti, A.; Smith, D. R. Science 2012, 337, 1072−1074. (27) Lassiter, J. B.; Chen, X.; Liu, X.; Ciracì, C.; Hoang, T. B.; Larouche, S.; Oh, S.-H.; Mikkelsen, M. H.; Smith, D. R. ACS Photonics 2014, 1, 1212−1217. (28) Im, H.; Bantz, K. C.; Lindquist, N. C.; Haynes, C. L.; Oh, S.-H. Nano Lett. 2010, 10, 2231−2236.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the National Science Foundation (NSF) CAREER Award (DBI 1054191) and ECCS 1610333. A.B. and X.S.C. acknowledge support from the University of Minnesota Doctoral Dissertation Fellowship. Device fabrication was performed at the University of Minnesota Nanofabrication Center, which receives support from the NSF through the National Nanotechnology Coordinated Infrastructure (NNCI) 6323

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324

Letter

Nano Letters (29) Chen, X.; Park, H.-R.; Pelton, M.; Piao, X.; Lindquist, N. C.; Im, H.; Kim, Y. J.; Ahn, J. S.; Ahn, K. J.; Park, N.; Kim, D.-S.; Oh, S.-H. Nat. Commun. 2013, 4, 2361. (30) Chen, X.; Park, H.-R.; Lindquist, N. C.; Shaver, J.; Pelton, M.; Oh, S.-H. Sci. Rep. 2014, 4, 6722. (31) Chen, X.; Ciracì, C.; Smith, D. R.; Oh, S.-H. Nano Lett. 2015, 15, 107−113. (32) Morgan, H.; Green, N. G. AC Electrokinetics: Colloids and Nanoparticles; Research Studies Press, Philadelphia, PA, 2003. (33) Dahlin, A. B.; Zahn, R.; Vörös, J. Nanoscale 2012, 4, 2339−2351. (34) Sbalzarini, I. F.; Koumoutsakos, P. J. Struct. Biol. 2005, 151, 182−195. (35) Squires, T. M. Lab Chip 2009, 9, 2477−2483. (36) Castellanos, A.; Ramos, A.; González, A.; Green, N. G.; Morgan, H. J. Phys. D: Appl. Phys. 2003, 36, 2584−2597. (37) Curto, A. G.; Volpe, G.; Taminiau, T. H.; Kreuzer, M. P.; Quidant, R.; van Hulst, N. F. Science 2010, 329, 930−933. (38) Pelton, M. Nat. Photonics 2015, 9, 427−435. (39) Kress, S. J. P.; Antolinez, F. V.; Richner, P.; Jayanti, S. V.; Kim, D. K.; Prins, F.; Riedinger, A.; Fischer, M. P. C.; Meyer, S.; McPeak, K. M.; Poulikakos, D.; Norris, D. J. Nano Lett. 2015, 15, 6267−6275. (40) Halas, N. J.; Lal, S.; Chang, W.-S.; Link, S.; Nordlander, P. Chem. Rev. 2011, 111, 3913−3961. (41) Duan, H.; Fernández-Domínguez, A. I.; Bosman, M.; Maier, S. A.; Yang, J. K. W. Nano Lett. 2012, 12, 1683−1689. (42) Kucsko, G.; Maurer, P. C.; Yao, N. Y.; Kubo, M.; Noh, H. J.; Lo, P. K.; Park, H.; Lukin, M. D. Nature 2013, 500, 54−58. (43) Rugar, D.; Mamin, H. J.; Sherwood, M. H.; Kim, M.; Rettner, C. T.; Ohno, K.; Awschalom, D. D. Nat. Nanotechnol. 2015, 10, 120−124. (44) Geiselmann, M.; Marty, R.; Renger, J.; Garcia de Abajo, F. J.; Quidant, R. Nano Lett. 2014, 14, 1520−1525. (45) Squires, T. M.; Messinger, R. J.; Manalis, S. R. Nat. Biotechnol. 2008, 26, 417−426. (46) Lal, S.; Link, S.; Halas, N. J. Nat. Photonics 2007, 1, 641−648. (47) Brolo, A. G. Nat. Photonics 2012, 6, 709−713. (48) Aouani, H.; Šípová, H.; Rahmani, M.; Navarro-Cia, M.; Hegnerová, K.; Homola, J.; Hong, M.; Maier, S. A. ACS Nano 2013, 7, 669−675.

6324

DOI: 10.1021/acs.nanolett.6b02690 Nano Lett. 2016, 16, 6317−6324